Master's Thesis completed 1998

Determination of Possible Wetland Mitigation Sites Using NC-CREWS and an Integer Linear Programming Formulation

Kevin Wayne Gainey

under the direction of Dr. Joseph Roise

Full text available here as an Adobe Acrobat PDF file

Roise, J.P., K.W. Gainey, T.H. Shear. 2004. An approach to optimal wetland mitigation using mathematical programming and geographic information system based wetland function estimation. Wetlands Ecology and Management 12(5):321-331.


We developed and tested a quantitative geographic information system (GIS)-based approach for selecting wetland restoration sites. Our approach uses a combination of an existing wetland function evaluation program, a GIS and integer programming methodology with an objective to minimize cost of restoration subject to meet environmental requirements. Investigations were conducted on the formulation to examine the effects of problem size, site ordering for input, and restoration targets. The formulation could be solved for the largest problem size tested of 996 integer variables. The larger the problem, the more time it took to solve. Larger restoration targets usually took more sites and more time to solve. Sorting sites by size was found to lead to inefficient and often unfeasible solutions. Random sorting of sites was found to be the more efficient method of inputting restoration sites into analysis.